re: Problem with syscall_disestablish() - PR kern/50430

[Paul Goyette <paul%vps1.whooppee.com@localhost>]

On Tue, 17 Nov 2015, matthew green wrote:

> Martin Husemann writes:
> > On Tue, Nov 17, 2015 at 05:48:17PM +1100, matthew green wrote:
> > > it also slows down normal operation adding branches and data access
> > > to the call path of every system call.  and bloats struct lwp and
> > > more.
> >
> > At least all operations on the syscall pathes need to be #ifdef MODULAR.
> >
> > How about just disabling all unloading for modules that ever used
> > syscall_establish() instead?
>
> ... or disable autounload by default anyway, because it's mostly a
> way to make my system slow.  i'm not sure i would go all the way
> and simply remove support for it completely, but i'd really much
> rather it wasn't triggered in a normal system.  the unload, and
> then reload, slow down the running of my apps, for no real benefit
> as an end user.  there is benefit for kernel developers, but those
> people can break the system in much more interesting ways, including
> asking for the unload manually when it's time to test something.

From sysctl(7) man page (manually re-wrapped for readability)

            kern.module.autotime
                     An integer that controls the delay before an
                     attempt is made to automatically unload a module
                     that was auto-loaded.  Setting this value to zero
                     disables the auto-unload function.

So just adding kern.module.autotime=0 in your /etc/sysctl.conf file
should disable auto-unload completely.


+------------------+--------------------------+-------------------------+
| Paul Goyette     | PGP Key fingerprint:     | E-mail addresses:       |
| (Retired)        | FA29 0E3B 35AF E8AE 6651 | paul at whooppee.com    |
| Kernel Developer | 0786 F758 55DE 53BA 7731 | pgoyette at netbsd.org  |
+------------------+--------------------------+-------------------------+